Date: Mon, 25 Oct 2004 00:52:25 +0200 From: Marius Strobl <marius@alchemy.franken.de> To: Kris Kennaway <kris@obsecurity.org> Cc: sparc64@freebsd.org Subject: Re: ntpdate unable to repair clock Message-ID: <20041025005225.A20538@newtrinity.zeist.de> In-Reply-To: <20041023231454.GA11787@xor.obsecurity.org>; from kris@obsecurity.org on Sat, Oct 23, 2004 at 04:14:54PM -0700 References: <20041023231454.GA11787@xor.obsecurity.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Oct 23, 2004 at 04:14:54PM -0700, Kris Kennaway wrote: > A sparc64 machine I'm using was consistently failing to set the date > properly when updated across the recent RELENG_5 changes that required > date to be reset. ntpdate claimed to be setting the date correctly, > but date(1) still returned a bogus value. > > Setting the date manually and then rerunning ntpdate gave another > bogus offset message, but in fact set it correctly. > You have to use `ntpdate -b` for such transitions so the time is also written to the clock chip. Your problems when not using the '-b' switch seem similar to what I also see on other architectures including i386 in this case. But there could be additional 64bit bugs in this path, e.g. the genclock clock isn't 64bit time_t clean, yet.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20041025005225.A20538>